Pattern-generating role for motoneurons in a rhythmically active neuronal network

Abstract

The role of motoneurons in central motor pattern generation was investigated in the feeding system of the pond snail Lymnaea stagnalis, an important invertebrate model of behavioral rhythm generation. The neuronal network responsible for the three-phase feeding motor program (fictive feeding) has been characterized extensively and divided into populations of central pattern generator (CPG) interneurons, modulatory interneurons, and motoneurons. A previous model of the feeding system considered that the motoneurons were passive followers of CPG interneuronal activity. Here we present new, detailed physiological evidence that motoneurons that innervate the musculature of the feeding apparatus have significant electrotonic motoneuron-->interneuron connections, mainly confined to cells active in the same phase of the feeding cycle (protraction, rasp, or swallow). This suggested that the motoneurons participate in rhythm generation. This was assessed by manipulating firing activity in the motoneurons during maintained fictive feeding rhythms. Experiments showed that motoneurons contribute to the maintenance and phase setting of the feeding rhythm and provide an efficient system for phase-locking muscle activity with central neural activity. These data indicate that the distinction between motoneurons and interneurons in a complex CNS network like that involved in snail feeding is no longer justified and that both cell types are important in motor pattern generation. This is a distributed type of organization likely to be a general characteristic of CNS circuitries that produce rhythmic motor behavior.

Fig. 1.

Motor and interneurons of the feeding system ofLymnaea in the “twisted” buccal ganglion preparation. A, The position of identified motor and interneurons that were the subjects of the present study. B1, B2, B3, B4, B4CL, B7a, B7b, B10, Motoneurons;N1M, N1 medial central pattern generator (CPG) interneuron; N1L, N1 lateral CPG interneuron;N2d, N2 dorsal CPG interneuron; N2v, N2 ventral CPG interneuron; N3p, N3 phasic CPG interneuron;N3t, N3 tonic CPG interneuron; SO, slow oscillator modulatory interneuron. Nerves of the buccal ganglia includeCBC, cerebrobuccal connective; BC, buccal commissure; DBN, dorsobuccal nerve; LBNand VBN, latero-/ventrobuccal nerve; andPBN, postbuccal nerve. A, Anterior;P, posterior; L, left; R, right. B, A summary diagram of the reciprocal synaptic connections between the modulatory SO cell and each of the feeding CPG interneurons, based on previous studies. Depolarization of the SO drives the CPG interneurons through these identified connections. The interneurons, in turn, produce the three-phase fictive feeding rhythm through complex synaptic connections (Elliott and Benjamin, 1985a) among N1, N2, and N3 phase cells (data not shown). Activity in the SO is entrained to the CPG rhythm by feedback from the interneurons.Black bars, Excitatory connections; black circles, inhibitory connections. The motoneurons (MNS) active in the three different phases of feeding (P, protraction; R, rasp;S, swallow) are driven by interneurons active in the same phase. C, Depolarizing current injected into the modulatory interneuron SO can drive a fictive feeding rhythm via sequential activation of the CPG interneurons. Ci, Maintained depolarization sufficient to excite the SO produces rhythmic burst activity in N1M, N2d, and N3p cells recorded at the same time.Cii, Expanded time base showing the characteristic waveforms of synaptic potentials and firing patterns of interneurons during N1/protraction, N2/rasp, and N3/swallow. The synaptic connections between the N cells allow the full N1–N2–N3 fictive feeding sequence to occur.

Fig. 2.

Electrotonic coupling between the protraction phase B7a motoneuron and the N1M CPG interneuron. A, Diagram showing the firing pattern of the B7a and N1M neurons in a CPG-driven fictive feeding rhythm. Both neurons fire together in the N1 (protraction) phase and are inhibited during the N2 and N3 phases.B, Depolarizing or hyperpolarizing current injected into B7a produces similar, although attenuated, voltage changes in N1M. Membrane potential (MP) changes in B7a were measured through a second voltage recording electrode so that coupling coefficients could be determined accurately. C, The B7a→N1M connection (Ci) is retained in HiLo+EGTA saline (Cii), which blocks chemical synapses and confirms the electrotonic nature of the connection. D, The N1M→B7a connection also involves an electrotonic synapse (Di), which persists in HiLo+EGTA saline (Dii). Note that in C andD all recordings were made through single current-passing electrodes, and so the presence of coupling could be established but could not be assessed quantitatively.

Fig. 3.

Dual role of the B7a as a motoneuron and the provider of an excitatory input to a CPG interneuron. A, Diagrammatic reconstruction of the morphology seen from joint fills of the B7a (5-CF, yellow; see Fig.4Ai) and the N1M neurons (MPTS,blue; see Fig. 4Ai). The B7a has a single axon projecting along the CBC and projecting from the latero- or ventrobuccal nerve roots. The N1M projects to the contralateral buccal ganglion and through to the contralateral CBC. The neuritic processes of these cells show extensive intermingling, and some of these are potential sites for electrotonic junctions (for high-magnification details of the areas in the rectangles, see Fig.4Aii, Aiii). B, B7a does not have direct connections with the B1 motoneuron but has indirect effects via the B7a→N1M→B1 pathway. Spikes evoked by depolarizing current injected into B7a excite N1M to firing through the electrotonic coupling between the two cells. Activation of N1M then evokes unitary EPSPs in the B1 through the N1M→B1 monosynaptic connection. In the absence of N1M spikes at the beginning of the trace, no inputs are seen on the B1. C, Dual function of B7a demonstrated by joint recordings from N1M and the posterior jugalis muscle (PJM). Activation of B7a evokes activity in the N1M interneuron via the electrotonic coupling. The B7a activation also causes a strong contraction of the protraction phase PJM buccal mass muscle recorded as voltage changes at the tip of a blunted electrode placed on the muscle surface. D, Activation of B7a can cause muscle contraction in the absence of spikes in N1M, indicating that the B7a→PJN connection is not mediated by other motoneurons driven by N1M.

Fig. 4.

Morphological demonstration of potential sites of electrophysiologically confirmed synaptic connections between motoneurons (injected with 5-CF, yellow) and interneurons (injected with MPTS, blue) in theLymnaea feeding system. Ai, Photomicrograph (magnified 200×, enlarged from a color slide taken through a 10× microscope objective) of a B7a motoneuron and an N1M interneuron. The main axon branches of the two cells run in close proximity in the buccal neuropile and cerebrobuccal connective. The two axons diverge at the branching point of the cerebrobuccal connective and latero-/ventrobuccal nerve. Aii, High-magnification (600×) photomicrograph (enlarged from a color slide taken through a 40× microscope objective with oil immersion) showing the area in therectangle in the left cerebrobuccal connective (LCBC) in Figure 3A. The depth of field with this objective and the filter set that has been used is ∼0.9 μm. The thickness of the branches shown is ∼4.0 μm; because the yellow (B7a) and blue (N1M) axons are both in the same (∼0.9 μm thick) plane of focus and appear to be in close contact (arrow), this provides a likely anatomical basis for the described electrotonic coupling between them to be direct.Aiii, High-magnification (600×) photomicrograph (enlarged from a color slide taken through a 40× microscope objective with oil immersion) showing the area in the rectangle in the left buccal ganglion in Figure 3A. Here, a branch of the initial axon segment of B7a makes a contact with the axon of N1M, making this another potential site of synaptic connections (arrow). B, High-magnification (400×) photomicrograph (enlarged from a color slide taken through a 20× microscope objective) showing the area in the rectanglein Figure 7D. Neurites from the motoneuron and interneuron are intermingled (arrow) in the region close to the cell bodies. Depth of field with 20× objective is ∼4.0 μm; the thickness of neurites shown is ∼2.5–4.0 μm. Ci, Photomicrograph (170×, enlarged from a color slide taken through a 20× microscope objective) showing the area in therectangle in the left buccal ganglion in Figure9E. Fine branches (∼2.0–4.0 μm) projecting from the axon of N3p are both in the same plane of focus and appear to be coming into close contact with the axon of a contralateral B4CL (arrowed area shown in inset, magnified 340×). Cii, Photomicrograph (170×, enlarged from a color slide taken through a 20× microscope objective) showing the area in the rectangle in the right buccal ganglion in Figure9E. Fine branches (∼3.0–4.0 μm) projecting from the initial axon segment of B4CL are both in the same plane of focus and appear to be coming into close contact (arrow) with the axon of the contralateral N3p (same pair of cells as inCi).

Fig. 5.

Rhythm-generating function of the B7a motoneuron.A, In this preparation the maintained depolarization of B7a could drive a fast fictive feeding rhythm. This occurs via strong activation of the N1M CPG interneuron. The modulatory interneuron, SO, is not activated by B7a but still receives subthreshold synaptic inputs from other CPG interneurons. B, In a more typical preparation, maintained depolarization can evoke one or two slow fictive feeding cycles. The B7a→N1M electrotonic synapse causes gradual depolarization of N1M, which eventually evokes a full fictive feeding cycle. C, Suppression of an SO-driven fictive feeding rhythm by hyperpolarization of B7a. Ci, In the control condition, maintained depolarization of the SO elicits a fast fictive feeding rhythm, which entrains the B7a. Cii, Maintained hyperpolarization of B7a prevents the same level of depolarizing current injected into the SO from driving a CPG rhythm. This presumably occurs because the hyperpolarized B7a suppresses N1M activity through its electrotonic connections and therefore prevents the SO from driving the CPG.

Fig. 6.

Resetting a spontaneous feeding rhythm by manipulating B7a spike activity. A, In an SO-driven feeding rhythm the N1M is driven into activity by the facilitating SO→N1M excitatory synaptic connection. This elicits individual synaptic potentials (arrow), which eventually trigger a burst of spikes in the N1M. Spike activity in B7a contributes to this buildup through the B7a→N1M electrotonic synapse. Hyperpolarization of the B7a slows this buildup and reduces the frequency of the pattern.B, Depolarization of B7a accelerates the buildup to plateau in the N1M, and the frequency of the rhythm is increased.C, Effects of briefer B7a activity changes on an ongoing SO-driven fictive feeding pattern. The noncoupled B7b also is recorded to monitor the effects on the rhythm. Ci, Hyperpolarization of B7a preventing one whole burst prolongs the duration of the cycle and delays the whole rhythm. Cii, Brief depolarization of B7a advances the next cycle by accelerating the onset of N2 inhibition on both SO and B7b. The B7a recording was made through a single current-passing electrode; therefore, the deflection in the trace does not reflect the actual size of the membrane potential shift. However, the depolarizing current was the same as in the experiments that were performed to demonstrate coupling between B7a and N1M, and the spike frequency in the evoked burst was in the range of spike frequencies observed in CPG-driven bursts. In both experiments the expected and actual N2 phases are marked before and after B7a manipulation. The vertical arrows show the phase shift that follows the perturbation (see Results for further explanation).

Fig. 7.

Electrotonic coupling between the rasp phase B10 motoneuron and the N2d/N2v retraction phase CPG interneurons.A, Diagram showing the firing pattern of the B10, N2v, and N2d neurons in a CPG-driven fictive feeding rhythm. All three neurons fire together in the N2 (rasp) phase. The N2 cells are inhibited during N3 and receive a biphasic input during the N1 phase. The B10 neurons are inhibited during N3 and are weakly excited during the N1 phase. B, Depolarizing or hyperpolarizing current injected into B10 produces similar, although attenuated, voltage changes in both the N2d and N2v. Membrane potential (MP) changes in B10 were measured through a second voltage recording electrode so that coupling coefficients could be determined accurately.C, The B10→N2d connection (Ci) is retained in HiLo+EGTA saline (Cii), which blocks (Figure legend continues.) chemical synapses, confirming the electrotonic nature of the connection. Note that the B10 recordings were made through a single current-passing electrode, and so the presence of coupling could be established but could not be assessed quantitatively. D, Diagrammatic reconstruction of joint fills of the B10 and N2d with 5-CF and MPTS, respectively. The B10 has a single axon that crosses the buccal neuropile and exits the ganglia via the postbuccal nerve. The N2d has axonal projections in both the postbuccal nerve and the ipsilateral dorsobuccal nerve. The extensive intermingling of neuritic processes from these two neurons suggests that the site of the electrotonic coupling is in the buccal neuropile (for high-magnification details of the area in the rectangle, see Fig. 4B). E, Depolarizing the B10 motoneuron occasionally can trigger spikes in both N2v and N2d CPG. In this example the activation of B10 depolarizes both N2-type interneurons via electrotonic coupling. This leads to the activation of a full plateau potential in N2v (see Brierley et al., 1997a) and spikes in the N2d, which are driven by both B10 and N2v cells. The electrotonically coupled network is shown in the schematic diagram.F, Motoneuronal function of B10 demonstrated by dual recordings from B10 and the muscle it innervates. Fi, Spike activity in B10, driven by a depolarizing stimulus, leads to the contraction of the radula tensor muscle (T).Fii, In the same preparation, during spontaneous fictive feeding, CPG-driven bursts of spikes in B10 also are followed by the contraction of the radula tensor muscle.

Fig. 8.

Resetting an ongoing SO-driven fictive feeding rhythm by manipulation of B10 activity. A, Hyperpolarization of B10 prevents the expected N2 phase of the feeding cycle and delays the rhythm. B, Depolarization of B10 has only a small effect on the ongoing rhythm, accelerating the buildup to N2d plateau and slightly advancing the next cycle. In both experiments the expected and actual N2 phases are marked before and after B10 manipulation. The vertical arrows show the phase shift that follows the perturbation.

Fig. 9.

Electrotonic coupling between the rasp/swallow phase B4CL and B4 motoneurons and the N3p and N3t CPG interneurons. A, Diagram showing the typical firing pattern of the B4CL, N3p, B4, and N3t neurons in a CPG-driven fictive feeding rhythm. The B4CL and N3p cells fire together at the end of the N2 (rasp) phase and continue firing into the N3 phase. The B4 and N3t cells fire together in the N3 (swallow) phase. All of the neurons are inhibited during N1 and the start of the N2 phase. B, Depolarizing or hyperpolarizing current injected into B4CL produces similar, although attenuated, subthreshold voltage changes in the N3p. The B4 motoneurons show a similar electrotonic connection with the N3t, but depolarizing current often can trigger full spikes in N3t. Membrane potential (MP) changes in B4CL and B4 were measured through a second voltage recording electrode so that coupling coefficients could be determined accurately. C, The B4CL→N3p connection seen in normal saline (Ci) is retained in HiLo+EGTA saline (Cii), which blocks chemical synapses, confirming the electrotonic nature of the connection. D, The N3p→B4CL connection is complex and probably consists of both chemical and electrotonic components. Di, Previous work has shown that the chemical component is inhibitory, and this is seen as the initial hyperpolarization on the B4CL. This is followed by a depolarization that reflects the conjoint electrotonic connection between them. This depolarization evoked axonal spikes in the B4CL motoneuron.Dii, The electrotonic component is revealed in HiLo+EGTA when a purely depolarizing potential follows N3p spike activity. The electrotonic synapse is also apparent when the N3p interneuron is hyperpolarized. Note that in C and D all recordings were made through single current-passing electrodes, and so the presence of coupling could be established but could not be assessed quantitatively. E, Diagrammatic reconstruction of the morphology seen from joint fills of the B4CL and N3p with 5-CF and MPTS, respectively. The B4CL projects into both ipsilateral and contralateral latero-/ventrobuccal nerves. The N3p projects into both the ipsilateral and contralateral dorsobuccal nerves. The neurites of these cells show considerable intermingling, and these are potential sites of electrotonic junctions (for high-magnification details of the areas in the rectangles, see Fig. 4Ci,Cii).

Fig. 10.

Dual role of the B4 cell as a motoneuron and a provider of inputs to CPG interneurons. Ai, Activation of B4 produces intense inhibition (expanded in Aii), which suppresses spontaneous bursting activity in the protraction phase CPG interneuron N1M. B, Evidence that this is likely to be attributable to the indirect B4→N3t→N1M pathway obtained from simultaneous intracellular recordings of B4, N1M, and N3t. Each spike in the N3t produces unitary IPSPs in the N1M. C, Hyperpolarization of B4 can release the N1M interneuron from the inhibition caused by tonic firing in N3t. A brief burst of spikes in N3t during this B4 suppression briefly interrupts firing activity in N1M. After the repolarization of B4 to resting potential, the N3t cell is released from inhibition; as a result, N1M spike activity is suppressed again. D, Simultaneous recording from the N1M, the swallow phase anterior jugalis muscle (AJM), and B4 shows that B4 has a dual function. A current-induced burst of spikes in B4 evokes unitary 1:1 excitatory junction potentials (EJPs) on the AJM and at the same time inhibits the N1M CPG cell. E, Summary of the synaptic connections of B4.

Fig. 11.

Resetting a feeding rhythm by manipulating B4 spike activity. Ai, Control experiment in which the N1M and B4 show bursting activity in the SO-driven fictive feeding rhythm.Aii, Faster time base and higher gain ofAi showing the normal level of N3t inputs appearing on the N1M during the buildup to the protraction phase. Bi, Hyperpolarization of the B4 advances the phase of the SO-driven rhythm. The N1M recovers more rapidly than normal and resets the subsequent rhythm. This is attributable to the reduced duration of the N3t inhibitory inputs on the N1M cell shown in detail inBii. Ci, Depolarization of the B4 during SO-driven fictive feeding delays the onset of the next feeding cycle and the subsequent fictive feeding rhythm. This is attributable to the increased duration of N3t inhibitory synaptic inputs shown at a higher gain and faster time base in Cii (see Results for further details). The B4 recording was made through a second voltage recording electrode so that the deflection in the trace reflects the actual value of the potential shift (12.5 mV). The direction of thearrows in the synaptic connectivity diagrams indicates the relative change in activity levels in the three neurons (up arrow, increase; down arrow, decrease).

Fig. 12.

Summary of electrotonic connections between motoneurons and CPG interneurons in the Lymnaea feeding system. Motoneurons are coupled only to interneurons that are active within the same phase of the feeding cycle. Only motoneurons that activate buccal musculature showed these connections. The protraction phase (P) motoneuron B7a is coupled to the CPG interneuron N1M. The rasp phase (R) motoneuron B10 is coupled to both the N2d and N2v CPG interneurons. N2d and N2v also are coupled electrotonically to each other, suggesting that these three neurons form a single coupled unit. The late rasp/swallow (R/S) phase motoneurons B4CL are coupled to the CPG interneuron N3p. The swallow phase (S) motoneuron B4 is coupled to the N3t interneuron. A cross-coupling also exists between the R/S and S neurons so that they form a larger coupled unit. Note that there are also chemical connections between the CPG interneurons and the motoneurons, which are not shown in this figure (for details of these, see Discussion).